中文摘要：

高质量的海底声强图是多波束海底底质分类的基础。在系统分析了多波束声强图像中沿测船航迹灰度异常值和区域性明暗差异成因的基础上,提出了基于加权最小二乘估计的多波束声强数据归一化方法。实例计算结果表明：该方法在不损失原有强度变化信息的基础上,有效地抑制了声强图中灰度异常值对海底底质分类的影响,提高了区域底质分类的质量。

英文摘要：

High-quality underwater sound image is the basis of seafloor classification by using sound images. After the problems of abnormal gray alongtrack in multibeam sound image were analyzed systematically, normalization of muitibeam echo intensity based on weighted least squares estimation method was proposed. The resuits showed that this method can effectively suppress the errors on seafloor classification derived of abnormal gray in multibeam sound image without loss of original echo intensity change information, and improved the quality of seafloor classification by using sound images.